The goal of this research is to develop a prototype tool that will ultimately improve cancer therapy. A major bottleneck in testing new cancer drugs is in the early phases of assessing drug activity, typically in phase II trials. The tool will combne computational methods of assessing cancer patients' tumor burden on CT images with new computational models of tumor growth over time. This systematic project will combine expertise among academic physician scientists in clinical pharmacology, oncology, and imaging with industry computational pharmacologists to develop a prototype tool for analyzing tumor burden and designing new, more efficient, clinical trials that could reduce the number of patients needed to test a new drug. This tool is also expected to enable investigators to better identify subsets of patients who are having greater benefits from treatment than others. Aim 1 entails computing the volume of tumors (rather than just single longest dimensions) for more than 900 patients each with colorectal, lung, and renal cancer and then establishing new longitudinal models of tumor growth based on the volumetric assessments. Aim 2 will project the earliest time points at which the tumor volume measurements detect treatment effects, simulate clinical trials based on the data in Aim 1 and validate the findings with prospective study of 90 cancer patients. Based on these findings, in Aim 3 the investigators will test the prototype tool by conducting prospective phase II trials with the new volumetric assessment and computational modeling-based study designs. As CT is the most common imaging modality for cancer, the new algorithms run on popular imaging platforms could then be readily implemented on a large scale.